Introducing Real Time Assist API

• Symbl.ai’s new Real-Time Assist API, powered by generative AI, provides real-time, dynamic guidance to agents during live conversations leveraging your business-specific knowledge base. It offers a comprehensive solution for live assistance, including objection handling, script adherence monitoring, compliance checks, and contextual responses to questions—all in real time.
• To learn more, explore our blog and documentation.

Introducing UI Components

• The UI Components are a set of pre-built, customizable UI designed to easily integrate Symbl.ai’s conversation insights into your application while aligning with your brand identity. Available as a React library, this version includes key components such as the meeting header, call score, and call sentiment, and can be tailored for various use cases.
• UI components allows users to breakdown each component such as Call Score, Sentiment and tailor each component to their preferences. UI components offers extensive customization options, enabling adjustments to color schemes, typography, and logos to match customer’s brand identity, and modifications to the layout with the ability to add or remove components. You can find UI components in our new npm library, and there’s also a Storybook available for you to explore and interact with the components.
• NPM Library: @symblai/symbl-ui-components
• Storybook: Symbl.ai UI Components Demo
• To learn more about UI Components, refer to docs.

New Features:

Custom Criteria for Call Score:

• Ability to create custom assessment criteria tailored to specific business goals such as empathy, technical clarity, and objection handling.

Scorecard for Call Score:

• Introduction of a customizable scorecard framework.
• Combination of user-defined and managed criteria for comprehensive performance evaluations.
  With this update, the Async API introduces an optional "scorecardId" parameter nested within the new "callScore" object. This parameter allows users to specify which scoring criteria should be applied by the Call Score API to each call.
• For more information, refer documentation:
  • Call score API: https://docs.symbl.ai/docs/call-score-api
  • Criteria: https://docs.symbl.ai/docs/criteria
  • Scorecard: https://docs.symbl.ai/docs/scorecard
  • Get started with call score: https://docs.symbl.ai/docs/get-started

Streaming API support for Call Score:

• No post-call processing of conversations needed
• Simple actions to auto trigger Call Score processing after the streaming session has ended
  To learn more, refer here.

Webhook for Call Score:

• Applicable for Async API and Streaming API
• Pushes notifications of status changes on user-specified URL
  To learn more, refer here.

Nebula Chat Model Playground

• Explore the Nebula Chat model with our new playground at https://nebula.symbl.ai/playground. Get started quickly with the sample prompts or create your own. The playground offers a no-code environment to try Nebula and iteratively fine-tune prompts for your use cases.
• Nebula is currently in private beta. Sign in or sign up on Symbl.ai platform to request access.
• To learn more, visit our docs.

Billing Page Improvements

• Invoice Management: Users can now view a comprehensive 10 billing cycle history of the invoices, amounts, and payment statuses, all from the billing page.
• Asynchronous Payment for Unpaid Invoices: Added a "Make Payment" button for unpaid invoices, streamlining the payment process and ensuring uninterrupted service.
• Buffer Period: Introduced a buffer period of 2 weeks post-payment failure, allowing users additional time to rectify payment issues without service disruption.

Subscribe API Enhancements

• A new event subscription_started is added to subscribe API when a subscription to a real-time conversation or stream is successfully initiated. The subscription_started event returns a conversation ID. Developers do not have to map the connection ID and conversation ID with this new event. For more information, see the Subscription Started Response Object docs

Go-SDK release

• The Symbl.ai Go SDK provides a convenient method to connect Symbl.ai conversation intelligence from applications written in Go. A prescribed or opinionated set of interfaces, classes, and functions help you quickly bootstrap to using the Symbl.ai platform.
• With Symbl.ai Go SDK, you can connect with our Streaming API (WebSocket enabled), Async API, Management APIs, and generate transcripts, conversation intelligence such as trackers, topics, action-items, entities, etc., and applied AI intelligence such as call scores and insights UI. For more information, see the Go SDK docs

Redaction for All Conversation APIs release

• The Redaction feature for all Conversation APIs, enhancing privacy and security in our services. This feature allows users to automatically redact sensitive information from conversation insights such as formatted transcripts, summaries, trackers, bookmarks, and more. Users can apply redaction on-demand with the redact=true parameter in API requests or set it as a default in their user settings, ensuring a flexible and secure conversational experience. For more information, see the Apply redaction to conversation APIs.

Nebula Embedding API release

• The Embedding API uses the Nebula embedding model to create vector embeddings from conversations or text data. Nebula's embedding model is optimized for calculating vector representations of conversation text (transcripts, emails, chats, etc.) while supporting the non-conversation text (documents, articles, etc.). For more information, see the Embedding API overview.

Nebula performance improvements

• The overall performance of Nebula has increased by +9% (R2), which improves its capability to generate more cohesive and insightful responses.
• Nebula can now process up to 8K tokens allowing it to process longer conversations.